Daniell on a Neio Pyrometer. 311 



it, and extending to 6 is a bar of platinum 10.2 inches long, 

 and 0.14 of an inch in diameter. It is immovably fixed at a 

 by a nut and screw of the same metal on the outside, and a pin 

 or shoulder on the inside. It is likewise confined to its place 

 at 6 by a small perforated plate of platinum through which it 

 passes. From its end b, proceeds a fine platinum wire of 

 about yjg of an inch diameter, and coming out of the tube at 

 d passes two or three times round the axis of the wheel i, fixed 

 on the back of the scale efg h, and represented at fig. 2. It 

 is then bent back and attached to the extremity of a slight 

 spring m n, which is stretched on the outside of the brass 

 ferule, and fixed by a pin at n. The wire is thus kept extended 

 by the action of the spring. The axis of the wheel i is 0.062 

 of an inch in diameter, ^and the wheel itself is toothed and plays 

 into the teeth of another smaller wheel k. This smaller wheel 

 is I the diameter of the larger, and carries on its circumference -^ 

 the number of teeth. To its axis, which passes through the 

 centre of tlie scale efg h, is attached the index I. Now the 

 theory of this combination is, that any alteration of the relative 

 lengths of the metal wires and earthen tube will cause the wheel 

 i to move from the action of the spring m n, which motion will 

 be multiplied three times by the wheel k, and measured by the 

 index I. The scale is divided into 360°. Instead of passing 

 the fine platinum wire round the axis of the wheel it has been 

 found better in practice to attach a short silken thread to its 

 extremity, and pass that round and fix it to the spring. The 

 dimensions, which I have stated above, may, of course, be 

 varied to suit difierent purposes. Nothing depends upon their 

 nice adjustment, or upon intricate calculation. The value of 

 the degrees, it will be seen in the sequel, is determined for 

 each instrument by two fixed points in a manner perfectly 

 analogous to the graduation of a thermometer. 



If the extremity of the instrument a 6 be now gently heated 



the index will be seen to move forward with a gradual and 



very equal motion, and by careful cooling, will return as 



gradually from the point from which it started. The accession 



Y 2 



